Qut Speaker Identity Verification System for Evalita
نویسندگان
چکیده
This document outlines the system submitted by the Speech and Audio Research Laboratory at the Queensland University of Technology (QUT) for the Speaker Identity Verification: Application task of EVALITA 2009. This competitive submission consisted of a score-level fusion of three component systems; a joint-factor analysis GMM system and two SVM systems using GLDS and GMM supervector kernels. Development evaluation and post-submission results are presented in this study, demonstrating the effectiveness of this fused system approach. This study highlights the challenges associated with system calibration from limited development data and that mismatch between training and testing conditions continues to be a major source of error in speaker verification technology.
منابع مشابه
The THUEE Speaker Identity Verification System for Evalita 2009 SIV-Application
This paper focuss speaker identity verification system from the Department of Electronic Engineering, Tsinghua University (THUEE) for Evalita 2009 Speaker Identity Verification Application Track (SIV-A). Two systems are submitted for the evaluation. The primary one consists of GMM-UBM and GMM-SVM. The contrast one is based on high level feature. We describe each subsystem briefly and give their...
متن کاملI3A System for Evalita 2009 Speaker Verification Application Evaluation
This paper describes the I3A-University of Zaragoza speaker recognition system in the Evalita 2009 Speaker Identity Verification Application task. Two systems have been submitted for evaluation: a primary system based on Joint Factor Analysis and a secondary system based on classical MAP adaptation without any channel compensation for comparison. We present results for both systems on the devel...
متن کاملAGNITIO’s Speaker Recognition System for EVALITA
AGNITIO’s submission for the EVALITA 2009 Speaker Identity Verification Application Track was a fusion of a state-of-the-art JFA system and a new I-Vector system. We briefly describe the two systems and their fusion and then highlight the challenging cross-channel conditions of this evaluation, where enrollment was via land-line and verification via mobile, or vice versa. Despite the availabili...
متن کاملThe QUT NIST 2004 Speaker Verification System: A fused acoustic and high-level approach
The trend towards including both acoustic and high level speech features in speaker recognition systems is addressed with the presentation of the speaker verification system developed by QUT for the NIST 2004 Speaker Recognition Evaluation. The system presented is a fusion of five subsystems including acoustic, lexical, phonetic, prosodic and durational speech features and focuses on utilising ...
متن کاملUsing Exciting and Spectral Envelope Information and Matrix Quantization for Improvement of the Speaker Verification Systems
Speaker verification from talking a few words of sentences has many applications. Many methods as DTW, HMM, VQ and MQ can be used for speaker verification. We applied MQ for its precise, reliable and robust performance with computational simplicity. We also used pitch frequency and log gain contour for further improvement of the system performance.
متن کامل